Why shell orientation is important?

[rubenvillarreal]

Hello,

I am doing some simulations using Creo. While doing simulations on Creo, is shell orientation important if we are using an ISOTROPIC material?

[SweetPeasHub · Accepted solution]

Creo Simulate does not really care about shell orientation. Flipping normals is not available on the simulate side - only the cad side.

Material orientation is handled separately although it may be less tedious to define material orientations with consistent normals. 

But, you will not need material orientations for isotropic and simple shells.

Convince yourself by flipping some normals in the CAD side.  (Pick surface, then editing dropdown, flip normal.)

rerun and see what happens.

In some other simulation software normals will be important, and this could change in CREO if they added or changed functionality.

[rubenvillarreal]

I think the same way. I believe that shell orientation can be important if you define some complex shells with layers etc..

I can change orientation in simulate (flipping arrows as CAD), however, a good question here is the following.

A co-worker uses Nastran to perform FEA. We used the same exact geometry, and regarding von Mises and displacement the results were almost the same (more than 95.6%).

However, our client says that we must include the top-bottom normal stresses plots in our report, and here we found that there is an enormous difference between his results and mines. I think that is because of shell orientation. Do you have any advice on this?

[SweetPeasHub]

I think it unlikely that you have the same vm stress but not the same stress tensors.

It is likely just how you are plotting the "normal stress".

I recommend to use max principal and min principal, top or bottom plots.

Be careful when selecting both top and bottom because one of the sides will not be the highest magnitude for that side.  The sign tension vs compression matters.  Max principal is the largest tension stress, min principal is the largest compression stress.  In other words you need two plots, one for compressive side and one for tensile.

The vector directions of the principal stresses can be shown in a vector plot in Creo.

In theory we should be able to get all the Cauchy stress tensor components and re-calculate the von-mises stress to confirm the calculation.

If you are still having concerns you may need to share more specific details about your model, mesh size control, boundary conditions, analysis type and setup, and specific results plots or measures needed.